rp8400/8420 Reliability Issues?

[Alzhy]

We've 3 rp84xx delivered and put on stream in the last 11 months. All 3 have failed hard despite compliant Data Center Conditions relating to clean power and climate control. The failures were:

1. System Board Failures (2 Frames)
2. CoreIO board failures (3 frames)
3. A failed fan on a CPU module that brought the entire frame down (1 occurence)

3 out of 3 critical failures out of an Enterprise Class server is the first I've seen amongst the big 3 UNIX vendors serving our enterprise. Could this just be coincidence or we're just unlucky on our part? Any sites out there have any experiences with rp84xx failures lately?

[Sunil Sharma_1 · Accepted solution]

It just coincidence i believe. We are using rp8400 since last two year and there is no hardware failure.

Sunil

[Jeff Schussele]

Hi Nelson,

We have a dozen or so 8400s here & have only seen 2 - 3 failures to date - none have TOC'd the box. But we've also found that the 8400 is very sensitive to firmware levels which is a pain because a FW update on it takes a good 2 1/2 hours because of all the cells that'll need it.
Frankly, we find them somewhat more reliable than our 'Domes.

My 2 cents,
Jeff

[Michael Tully]

We have four of these and have had only one failure. That was near on 3 years ago when we had a failed fan. Other than that, we have a reasonable run, except where we have done patch updates or logically moving cell boards, where the system gets stuck in the BIB (boot is blocked).

my aussie 3 cents

[Tim D Fulford]

We delivered 8 rp8400 servers with 750MHz CPUs to some custoers and had no problems. Then we delivered about 8 rp8400 with 875MHz CPUs to one customer and 3 of them had failed system boards (or CoreIO) and 2 CPU failures etc before they were commissioned & one had a failed CPU after it was commissioned.

My suspision was the 875 flavor and I dont think they quite reach the standard.... but this is only a suspicion.

Regards

Tim

[Michael Tully]

Just for the record, our rp8400's are all 750MHz.